The purpose of this study was to compare the effects of multiage and homogeneous age methods of grouping pupils for classroom instruction in the primary department of Beaverbrook Elementary School in Griffin, Georgia. Results were expected to aid the staff in developing a more comprehensive and functional nongraded school. The controlled variable was the method used in grouping pupils. Six homogeneous age and seven multiage classes were established with pupils in their first, second and third years of school. Both groups were pretested and posttested in reading and mathematics achievement and self-concept. First-year pupils' achievement gain was significant at the 0.03 level in favor of the multiage group. Pupils in their second and third years of school showed no significant gain in reading achievement; however, the mathematics gain favored the homogeneous age group and was significant at the 0.02 level. Self-concept gain was significant at the 0.02 level in favor of the multiage group. The data were computed from a random sample of 120 pupils, approximately one-third of the population. A comparison of the two methods of grouping indicated the multiage method to be more appropriate for primary students in the school. A review of selected literature on grouping is included in the document. (Author/MS)